Tablet-Based Method for Handwriting Assessment

Background:  Handwriting is crucial for success in school, daily living, and communication with others. Difficulties with handwriting can interfere with a child’s ability to dedicate cognitive resources necessary to learn a wide range of subject matter and consequently mask a child’s capabilities in other areas. As high as 37% of children are estimated to have handwriting difficulties; it is one of the most common reasons children are referred for occupational therapy and is associated with a number of developmental diagnoses –including children with autism spectrum disorder (ASD) who struggle to develop a wide range of skilled motor behaviors and are therefore particularly vulnerable to handwriting impairments. Existing handwriting assessments are limited to labor-intensive manual measures of letter formation that produce qualitative, categorical (correct vs. incorrect) ratings and computerized kinematic analyses that lack any measures of letter-form.

Objectives:  To developed a fully automated cloud-based handwriting analysis program for Android and Apple tablets that seamlessly enables evaluators and interventionists across a range of disciplines to quantify handwriting performance across multiple dimensions.

Methods:  Our approach incorporates both real-time kinematic feedback as well as a detailed, quantitative measure of letter formation. This novel and unique letter-form analysis employs a computational morphometric algorithm that maps the template letter to the subject’s letter, and in so doing, generates a continuous measure of letter formation that is highly reliable. To date, a research-suitable version of this program has been developed and piloted in children using two tasks: English characters (42 Typically Developing [TD], 6F; 28 ASD, 2F; 23 Attention-Deficit Hyperactivity Disorder [ADHD], 4F) and Non-English characters (45 TD, 7F; 25 ASD, 3F; 25 ADHD, 5F). For each task, children completed three conditions: copy, trace, and fast trace. All groups were balanced for age, sex, PRI (WISC-IV), Socio-economic status, and handedness. Motor abilities were assessed using the Physical and Neurological Examination for Subtle Signs (PANESS).

Results:  Consistent with previous research, our novel computerized assessment of letter-form  revealed impairments in children with ASD compared with TD children across all tasks and conditions (English copy p =.031, trace p<.001, fast trace p<.001; Non-English copy p =.062, trace p=.014, fast trace p=.021; figure 1), as well as significant differences between ASD and ADHD in the English fast trace condition (p=.036; figure 1). No group differences (TD, ASD, and ADHD) were observed in terms of handwriting kinematics.

Letter form, in both tasks, was correlated with the WISC-IV’s working memory index across all conditions in the ASD group (p<.008) and with PANESS total score for the trace condition (p<.05). These results may suggest decreased automaticity and greater recruitment of higher order cognitive systems (e.g. mPFC and DLPFC) in the ASD group. 

Conclusions:  Our novel method is sensitive to diagnostic differences, particularly in ASD and can be used for both English and Non-English characters which allows for inclusion of an international community. The tablet based handwriting assessment will empower both professionals and interventionists to readily implement and evaluate the efficacy of targeted interventions for handwriting, serving a large public health need.